As the number of breast cancer (BCa) survivors increases, medication adherence has been identified as a key challenge in their long-term treatment. Current guidelines recommend that postmenopausal women with hormone receptor-positive BCa take aromatase inhibitors (AI) to improve survival. However, a large subgroup of BCa survivors discontinues AIs prematurely (24%), and up to 50% of survivors poorly adhere to AIs, potentially putting their lives at risk. A primary reason for AI non-adherence is musculoskeletal pain (MSP). Up to 61% of BCa survivors taking AIs develop MSP. MSP can be grueling, including severe joint, bone, and muscle pain and stiffness particularly in the wrists, hands and knees. MSP can lead to an inability to perform daily activities and can reduce quality of life in BCa survivors. Hypnosis, a mind-body intervention, has been widely demonstrated to reduce pain. In fact, analgesia is perhaps the best known effect of hypnosis. Meta- analyses have consistently supported the efficacy of hypnosis for pain control, including chronic pain. The results of our group's randomized controlled trial (RCT) using hypnosis for pain control in BCa surgical patients were consistent with published effects. The cost-effectiveness of hypnosis in BCa settings has also been supported (e.g., average cost savings of $772.71/patient). What is known is that hypnosis has been shown to be efficacious in managing pain, and that pain (MSP) is associated with AI non-adherence. What is unknown is whether hypnosis, by controlling MSP, can improve AI adherence (proposed mediational model). The study will directly compare two methods of delivering hypnosis [live hypnosis (H-L), app hypnosis (H-App)] on two sets of outcomes, clinical efficacy (decreased MSP, increased AI adherence) and cost-effectiveness. We felt that a hypnosis app was important to test as a potentially inexpensive and scalable means of hypnosis delivery. The Specific Aims of this RCT are: Aim 1) To investigate the relative clinical efficacy and cost-effectiveness of two methods (H-L vs. H-App) of delivering a hypnosis intervention to reduce MSP in BCa survivors taking AIs. Patients (N=360) from the Mount Sinai Medical Center will be randomly assigned to 1 of 3 groups: H-L, a live clinician-led hypnosis intervention; H-App, an app-based hypnosis intervention; and, AC, a live attention control group (empathic listening). Aim 2) To investigate the relative clinical efficacy and cost-effectiveness of two methods (H-L vs. H-App) of delivering a hypnosis intervention to increase AI adherence in BCa survivors. Aim 3) To test the hypothesized mediational model: Hypnosis exerts its effects on AI adherence through MSP reduction. This mediational model will be tested using a latent growth curve modelling (LGCM) approach and will help us understand the underlying mechanisms responsible for hypnosis effects.